Data and code from: Telomere length is determined by intrinsic factors and is shortened during drought years in Gallotia galloti
Data files
Apr 30, 2026 version files 8.08 MB
-
CT_values_All.csv
32.47 KB
-
CT_values_v2.csv
43.06 KB
-
Raw_plate_read_data.zip
7.91 MB
-
README.md
8.87 KB
-
telomere_length_Ggalloti_script_EG_etal.R
23.74 KB
-
TL_temporal_transformed_hw_dr.csv
30.95 KB
-
TL_temporal_transformed.csv
29.65 KB
Abstract
This repository contains the raw data and annotated R code associated with the study of relative telomere length (rTL) dynamics in the Western-Canaries Lizard (Gallotia galloti) across Tenerife, Canary Islands, Spain. The dataset comprises qPCR-derived relative telomere length measurements from lizard tissue samples (blood and tail tip) collected across multiple localities spanning an elevational gradient, sampled across several years (2017–2022). Raw Cq values from qPCR plates are included alongside derived rTL values calculated using the Pfaffl method, which accounts for plate-specific amplification efficiencies. Additional biological and environmental variables accompany each sample record, including sex, snout-vent length, morphotype, elevation, and microclimate data (relative humidity, wind speed, solar radiation, radiant sky temperature).
The R code covers the full analytical pipeline: (1) calculation of relative telomere length from raw CT values and assessment of inter-plate repeatability via intraclass correlation coefficients; (2) principal component analysis to reduce correlated environmental variables; (3) generalised linear mixed-effects modelling (GLMM) using a Gamma distribution with year as a random effect; (4) multi-model inference via AICc-based dredging and model averaging to identify key predictors of rTL; (5) hierarchical variance partitioning; and (6) sensitivity analyses examining the influence of tissue type, locality, and morphotype on model outcomes.
Dataset DOI: 10.5061/dryad.4f4qrfjt7
Description of the data and file structure
README: Data and Code Repository
Title: Telomere length is determined by intrinsic factors and is shortened during drought years in Gallotia galloti
Authors: Edward Gilbert, Megan L. Power, Annika Wolberg, Rodrigo Megía-Palma, Anamarija Žagar, Marta López-Darias, Miguel A. Carretero, Nina Serén, Pedro Beltran-Alvarez, Katharina C. Wollenberg Valero.
Description
This repository contains the raw qPCR data, processed datasets, and R code used to investigate drivers of relative telomere length (rTL) in the Western-Canaries Lizard (Gallotia galloti) across Tenerife, Canary Islands, Spain. Data were collected across multiple localities spanning an elevational gradient between 2013 and 2022.
File Descriptions
CT_values_All.csv — Raw Cq values from all qPCR plates used to calculate relative telomere length via the Pfaffl method.
CT_values_v2.csv — Raw Cq values, including Gold standard reference samples, used for inter-plate repeatability (ICC) assessment.
TL_temporal_transformed.csv — Main analytical dataset containing derived rTL values alongside biological (sex, SVL, morphotype, tissue type) and environmental (elevation, microclimate) variables for each individual.
TL_temporal_transformed_hw_dr.csv — As above, with additional columns for annual heatwave and drought frequency used in cross-correlation analyses.
Complete_tidied_telomere_script.R — Full R analysis pipeline including rTL calculation, PCA, GLMM modelling, model averaging, hierarchical partitioning, sensitivity analyses, and cross-correlation with climate data.
Raw_plate_read_data.zip — Raw output files from six qPCR plate runs (ThermoFisher QuantStudio Flex system) prior to data processing.
Software
All analyses were conducted in R (v4.5.1). Key packages: glmmTMB, MuMIn, hier.part, glmulti, emmeans, performance, effectsize, rptR, ggplot2, dplyr.
Licence
This dataset is made available under a Creative Commons Attribution (CC0) licence waiver.
Files and variables
File: telomere_length_Ggalloti_script_EG_etal.R
Description: R script containing the full analysis pipeline, including relative telomere length calculation from raw Cq values (Pfaffl method), inter-plate repeatability (ICC), principal component analysis of environmental variables, generalised linear mixed-effects modelling (GLMM) with AICc-based dredging and model averaging, hierarchical variance partitioning, exhaustive GLM model selection (glmulti), sensitivity analyses (tissue type, locality, morphotype, known-sex subsets), standardised effect size estimation, emmeans post-hoc comparisons, and cross-correlation analysis of rTL with annual drought and heatwave frequency.
File: TL_temporal_transformed.csv
Description: Main analytical dataset containing derived rTL values alongside biological (sex, SVL, morphotype, tissue type) and environmental (elevation, microclimate) variables for each individual.
Variables
- Year: Calendar year in which the individual was sampled.
- SampleName: Unique identifier for each sampled individual, corresponding to field and laboratory records.
- Env.type: Broad environmental category of the sampling locality based on elevation and climatic conditions.
Categorical variable (e.g., A = lowland coastal, B = mid-elevation, C = humid coastal, D = high elevation). - locality_code: Abbreviated code identifying each sampling locality.
- locality: Full name of the sampling locality.
- elevation: Elevation of the sampling locality above sea level (metres, m).
- max.temp: Mean annual maximum air temperature at the locality (degrees Celsius, °C).
- mean.temp: Mean annual air temperature at the locality (degrees Celsius, °C).
- min.temp: Mean annual minimum air temperature at the locality (degrees Celsius, °C).
- RH.mean: Mean annual relative humidity at the locality (percentage, %).
- wind.mean: Mean annual wind speed at the locality (metres per second, m·s⁻¹).
- sol_rad: Mean annual surface solar radiation (watts per square metre, W·m⁻²).
- TSKY.mean: Mean annual sky (radiative) temperature derived from microclimate models (degrees Celsius, °C).
- sex: Sex of the individual:
F = female, M = male, U = sex undetermined. - SVL: Snout–vent length, a proxy for body size and age (centimetres, cm).
- morphotype: Morphological subspecies definition within Gallotia galloti (e.g., galloti, eisentrauti, insulanagae).
- tissue_type: Tissue used for DNA extraction and telomere analysis (e.g, blood, tail).
- zscore: Z-score–standardised value of log-transformed relative telomere length (dimensionless).
- ratio: Raw relative telomere length (telomere repeat copy number to single-copy gene copy number ratio; T/S ratio), dimensionless.
- Age: Estimated age of the individual (years), derived from growth models based on SVL.
- Age2: Quadratic age term (Age²), included to model non-linear relationships between age and telomere length.
File: TL_temporal_transformed_hw_dr.csv
Description: As above, with additional columns for annual heatwave and drought frequency used in cross-correlation analyses. Note: Missing values are intentionally retained to preserve continuous annual climate records even in years without biological sampling.
Variables
- All variables listed for
TL_temporal_transformed.csv, plus: - Total_Heatwaves: Total number of heatwave events recorded in Tenerife during the given year (count).
- Total_Droughts: Total number of drought events recorded in Tenerife during the given year (count).
File: CT_values_All.csv
Description: Raw Cq values from all qPCR plates used to calculate relative telomere length via the Pfaffl method.
Variables
- Sample: Sample identifier.
- scg: Quantification cycle (Cq) value for the single-copy reference gene.
- tel: Quantification cycle (Cq) value for the telomere repeat amplification.
- eff.scg: Amplification efficiency of the single-copy gene reaction (unitless).
- eff.tel: Amplification efficiency of the telomere reaction (unitless).
- plate: Numeric identifier of the qPCR plate on which the sample was run.
- ref.tel: Mean telomere Cq value of the reference (gold standard) sample on the same plate.
- ref.scg: Mean single-copy gene Cq value of the reference (gold standard) sample on the same plate.
File: CT_values_v2.csv
Description: Raw Cq values, including Gold standard reference samples, used for inter-plate repeatability (ICC) assessment.
Variables
- Sample: Sample identifier, including biological samples, gold standard controls (“Gold”), and negative controls (“Neg”).
- scg: Quantification cycle (Cq) value for the single-copy gene.
- tel: Quantification cycle (Cq) value for the telomere repeat.
- eff.scg: Amplification efficiency of the single-copy gene reaction (unitless).
- eff.tel: Amplification efficiency of the telomere reaction (unitless).
- plate: Identifier for the qPCR plate used.
- ref.tel: Reference telomere Cq value used for plate normalisation.
- ref.scg: Reference single-copy gene Cq value used for plate normalisation.
File: Raw_plate_read_data.zip
Description: Compressed archive containing raw output files from six qPCR plate runs performed on a ThermoFisher QuantStudio Flex system. These files represent the instrument-level data prior to any processing or rTL calculation.
Contents: 6 × qPCR plate output files (.csv)
Code/software
RStudio v.5.3.1 and packaged in detail within the R script, shared here as follows:
Modelling
glmmTMB— generalised linear mixed-effects models (Gamma GLMM)MuMIn— multi-model inference, AICc-based dredging and model averagingglmulti— exhaustive GLM model selection (alternative approach)rptR— intraclass correlation coefficients (inter-plate repeatability)
Model evaluation & effect sizes
performance— pseudo-R² for GLMMs, ICCeffectsize— standardised effect size estimationr2glmm— R² for mixed-effects modelsemmeans— estimated marginal means and post-hoc pairwise comparisonshier.part— hierarchical variance partitioning
Data manipulation
dplyr— data wrangling and summarisationtidyr— data reshaping
Visualisation
ggplot2— all figuresviridis— colour palettes
Access information
Other publicly accessible locations of the data:
Relative telomere length (rTL) was measured in Gallotia galloti tissue samples (blood and tail tip) collected across multiple localities on Tenerife, Canary Islands, Spain, spanning an elevational gradient from sea level to the summit of Mount Teide. Samples were collected across multiple years (2017–2022) from individuals representing two morphotypes (galloti and eisentrauti, and 6 samples from the rarely sampled G. g. insulanagae). Telomere length was quantified using quantitative PCR (qPCR), with rTL calculated from raw Cq values following the Pfaffl (2001) method, which corrects for plate-specific amplification efficiencies of both the telomere primer pair and the single-copy gene (SCG) reference. Triplicate reactions were averaged per sample. Inter-plate consistency was assessed using intraclass correlation coefficients (ICC) estimated with restricted maximum likelihood via the rptR package in R.
Environmental variables (elevation, relative humidity, wind speed, solar radiation, and radiant sky temperature) were reduced to orthogonal axes using principal component analysis (PCA) prior to modelling. Biological variables recorded per individual included sex, snout-vent length (SVL), morphotype, and tissue type. All analyses were conducted in R (v4.5.1).
To identify predictors of rTL, a Gamma-distributed generalised linear mixed-effects model (GLMM) was fitted using glmmTMB, with year as a random effect to account for temporal pseudoreplication. Fixed effects included environmental PC axes, SVL, sex, morphotype, tissue type, and environment type. Multi-model inference was performed by exhaustive dredging of the global model using MuMIn, retaining all models within ΔAICc < 2 for model averaging, which yielded weighted parameter estimates and unconditional confidence intervals. The relative importance of predictors was further assessed using hierarchical variance partitioning. Sensitivity analyses were conducted to evaluate the robustness of results to tissue type (blood-only and tail-only subsets), inclusion of locality as an additional random effect, and morphotype composition. As a complementary approach, an exhaustive GLM selection algorithm was implemented using the glmulti package, fitting a Gamma GLM (log link) across all possible combinations of predictors, including: year, environment type, raw climatic variables, sex, SVL, morphotype, and tissue type - ranked by AICc. Predictor importance was validated by examining variable occurrence frequency across the top 100 candidate models. Post-hoc pairwise comparisons were conducted using emmeans. This GLM-based approach served as a cross-check of the GLMM dredging results. Temporal patterns in rTL were additionally examined in relation to annual climate anomalies using cross-correlation coefficient analysis, verified by permutation testing, to assess the influence of extreme weather events across the sampling period.